Daniel Price, head of social operations at social and content agency Lost Boys, looks at the power of social images, and how brands can harness it.
Trillions of conversations, both public and private, buzz across the social web every day. And these conversations are being listened to, and not just by the social channels themselves. Brands are listening, and using what they hear to their advantage.
More and more brands are analysing this social media data to measure the impact of their campaigns and monitor mentions of their brand, products and competitors, with a growing number of tools available to help them do it. These tools promise to help brands derive insights from the social noise, and they’re becoming an essential part of the marketer’s toolkit.
Traditionally, social listening has been fraught with difficulties: inconsistency of data across different channels, measuring intangibles like sentiment, over-reliance on Twitter data because it’s readily available. The list goes on. But the more fundamental problem with social listening is its focus on analysing keywords and text associated with posts on social channels.
Currently, the majority of social analysis doesn’t take into account the 1.8bn images posted to social every day and (more importantly) the meanings held within these images. After all, a picture is worth a thousand words. However, this is all about to change with some interesting developments in social tools, which will soon allow us to search within images.
Facebook and Google have already developed ways to recognise your friends in photographs and even reverse image search. Now, social listening platforms are starting to use image recognition technology in order to help brands harness this rich data source.
Humans process images more quickly than words, and campaigns featuring images tend to deliver higher levels of engagement. It’s not surprising that photo-based networks such as Instagram, Tumblr and Vine are flying while Twitter is seeing growth stagnate. Brands are now investing large sums in Instagram, hoping to leverage influencers in order to generate organic reach, but it’s tricky to measure the impact of this. This has pushed brands to focus on paid content on Instagram, which of course comes with its own set of metrics.
A number of social listening platforms, such as Synthesio, Brandwatch, Sysomos and Pulsar, are beginning to launch tools which analyse the content of images – recognising brand logos, text on images and even faces and facial expressions. These tools have the potential to unlock some of the most popular social network s to more meaningful analysis, particularly important as it’s claimed that 85 per cent of social photos featuring brands don’t include a mention of the brand in their accompanying text.
So what insights can we glean from images found on the social web? As with all things, they’re meaningless without context. Pairing up this new data stream with geo data and text analysis can start to give us a much more rounded picture of how people are interacting with your brand and your products in the real world.
And that’s not all. We can also start to crack some of the other issues we have with social data: notably, the issue of sentiment analysis. Recently, some more sophisticated tools have added smile recognition analysis into the mix to help determine how people feel about brands. While we’re some way from telling the difference between a smirk and a smile, the journey has begun.